Client profitability analysis compares the revenue from a client with the resources used to serve that client during the same period. Start with fees, logged delivery time and attributable costs. Show shared overhead separately, explain the allocation rule and keep payment timing in its own record. The result should lead to a specific conversation about scope, staffing or price.
The fictional example below compares three agency clients. Its largest account generates the least delivery contribution. That finding is a reason to inspect the work, not an automatic instruction to dismiss the client. All amounts and rates are illustrative USD planning assumptions.
Define what belongs in the review
Choose a month and use the same revenue policy for every client. For a retainer, record the service fee assigned to the period. For a project, check the milestones or completion rule used in your accounts. Do not compare a project's entire contract value with only one week of its costs.
Create a client ID and project ID so time and expenses reach the right account. Include discovery, delivery, account management, revisions and reporting when they are part of serving that client. Keep internal selling time visible in the appropriate agency category instead of quietly attaching it to whichever client has the biggest fee.
Exclude client-funded media from this management example. If your real review includes pass-through expenditure, retain a separate fee view and ask your accountant how the arrangement affects reported revenue. Document refunds, credits and disputed work. An unpaid invoice may require accounting review, but it is not automatically equivalent to extra delivery hours.
The CFI customer-profitability overview describes the purpose of evaluating profitability by customer. This article's tables use original fictional numbers and an explicit agency cost policy.
Use a loaded rate that you can explain
A person's selling rate is what the client pays; a loaded cost rate estimates the employment cost assigned to an hour of work. They serve different purposes. The rate policy needs to identify included compensation and employment costs, the period and the hours used as its denominator.
For illustration, $6,000 of monthly employment cost divided by 120 planned delivery hours produces $50 per delivery hour. That is one planning method. If the same cost is divided by 160 paid hours, the result is $37.50, with non-delivery time allocated elsewhere. Either method requires a reconciliation to the agency total. Do not mix them or add the same payroll again as overhead.
For owner time, keep an explicit planning allowance if the accounts do not already include the appropriate compensation. For contractors, check whether their invoice is already included in an hourly allocation before adding it as an outside cost. The goal is complete costs without double counting.
Three clients, one month
Juniper Studio reviews active clients Alder, Birch and Cedar. In this simplified example, all delivery work uses the same $50 loaded planning rate. Client-specific tools and contractors are added separately. The studio has $1,800 of remaining shared overhead that is not included in those rates or direct costs.
| Client | Service revenue | Delivery hours | Labor at $50/hour | Tools | Contractors | Delivery cost |
|---|---|---|---|---|---|---|
| Alder | $7,000 | 100 | $5,000 | $500 | $1,000 | $6,500 |
| Birch | $4,000 | 35 | $1,750 | $200 | $400 | $2,350 |
| Cedar | $2,500 | 20 | $1,000 | $100 | $0 | $1,100 |
| Total | $13,500 | 155 | $7,750 | $800 | $1,400 | $9,950 |
Delivery contribution equals service revenue minus the defined delivery cost. Effective revenue per delivery hour equals service revenue divided by logged delivery hours. It is a workload indicator; it is not a wage rate or a full profit measure.
| Client | Delivery contribution | Contribution margin | Revenue per delivery hour |
|---|---|---|---|
| Alder | $500 | 7.14% | $70.00 |
| Birch | $1,650 | 41.25% | $114.29 |
| Cedar | $1,400 | 56.00% | $125.00 |
| Portfolio | $3,550 | 26.30% | $87.10 |
The portfolio margin is $3,550 / $13,500 = 26.30%. Averaging 7.14%, 41.25% and 56% equally would ignore the different revenue sizes. Portfolio revenue per hour is $13,500 / 155 = $87.10, not the unweighted average of the three rates.
Alder contributes fewer dollars than either smaller account. Review where its 100 hours went before deciding what to change. A complex launch month could explain a temporary result; repeated unapproved revisions could point to an ongoing scope problem.
Show overhead without pretending the allocation is exact
For this review, the owner assigns the $1,800 remaining overhead in proportion to service revenue. Alder receives 7,000/13,500 of the total, Birch 4,000/13,500 and Cedar 2,500/13,500. The table rounds to cents and puts the one-cent rounding adjustment in Cedar's allocation.
| Client | Revenue-based overhead allocation | Result after allocated overhead |
|---|---|---|
| Alder | $933.33 | −$433.33 |
| Birch | $533.33 | $1,116.67 |
| Cedar | $333.34 | $1,066.66 |
| Total | $1,800.00 | $1,750.00 |
This is an allocation policy, not proof that Alder caused exactly $933.33 of overhead. Other policies, such as relevant support hours, can produce different client results. Explain the basis and use it consistently. Do not label the subtotal accounting net profit when financing, tax or other costs remain outside it.
An allocated loss does not establish that ending Alder would increase total profit. Some shared costs may remain, and the released capacity may not be sold immediately. Keep delivery contribution, avoidable expenses and allocated overhead as separate decision inputs. Use the agency-wide margin example to reconcile the whole business view.
Turn Alder's result into a workable decision
The team checks the time log and finds 15 hours of extra revisions beyond the expected workflow. At $50 per hour, that is $750 of allocated delivery effort. If next month's equivalent work genuinely takes 85 hours with the other costs unchanged, delivery cost becomes $5,750 and contribution becomes $1,250, or 17.86% of the same $7,000 fee.
That improvement is a scenario. The team first needs to distinguish mistakes it should correct from new client requests. It can then agree on deliverables, review rounds and approval checkpoints. The discussion should concern the actual work and client goals, rather than blaming the client for a low spreadsheet score.
There is also a price scenario. At the current $6,500 delivery cost, a 25% delivery contribution target would require $6,500 / 0.75 = $8,666.67. The arithmetic does not establish a market price or a recommended increase. The owner must test scope, value, buyer willingness and the existing agreement. A smaller scoped offer may fit better than a higher fee for the same open-ended work.
Keep capacity and collections beside the margin
Birch's healthy contribution does not mean the studio can accept unlimited similar work. Check the required roles and weeks against the agency capacity plan. If the 35 hours are concentrated in a week where the designer is fully committed, the monthly average hides a delivery conflict.
Collections need a separate check. If Alder's $7,000 invoice remains unpaid at month end, the service revenue and delivery effort may still appear in the profitability review, while available cash is lower. Record expected collection dates in the cash-flow forecast. Do not add the invoice to both opening cash and future receipts.
Strategic importance deserves a written note, not an unlimited exemption. A new account may be a deliberate test of a service. Record the test budget, expected learning and review date so an experimental month does not become a permanent unexplained loss.
Copy a monthly client-review register
| Field | What to record |
|---|---|
| Period and client/project ID | The month and exact account being reviewed |
| Service revenue and recognition basis | Fee, project milestone and relevant credit |
| Time by role and stage | Delivery, management, revisions and approved changes |
| Cost-rate version | Included costs, denominator and effective date |
| Direct expenses | Client tools, contractors and other attributable costs |
| Contribution and overhead policy | Both subtotals with their definitions |
| Collections and next due date | Separate cash evidence |
| Explanation and next action | What changed, owner and review date |
Review missing time before ranking clients. A zero-hour account might reflect incomplete records rather than remarkable efficiency. Inspect duplicate time entries, mismatched dates and contractor invoices attached to the wrong project. Keep corrections traceable and rerun totals.
After the review, select one action per material issue: clarify scope, repair an estimate, inspect the staffing mix or discuss a renewal. Check the next comparable period. A pricing change that loses suitable work can have a different effect than the spreadsheet predicted.
Using an offline client-planning tool
Agency Client Profitability & Capacity OS describes client records, time and costs, loaded team rates and capacity planning. Its guide explains that historical retainer views use current client settings, so keep dated evidence if a past fee or allocation changes. CSV exchange covers time entries; full-workspace backup uses JSON. It does not replace an accounting ledger or automatically synchronize across devices.
Use fictional data in the public preview and test a small case you can calculate independently. Check whether its revenue and allocation rules match your review. A tool is useful when you understand the inputs and can explain the report to the people making the decision.
Questions before acting on the result
Should I drop the least profitable client?
First verify the record and distinguish contribution from allocated profit. Examine avoidable costs, capacity, collection risk and the work that could replace the account. A ranking alone cannot answer the decision.
What if time tracking is incomplete?
Label the result incomplete, gather a short measured period and use clearly labeled estimates only where necessary. Do not give unrecorded time a value of zero without explaining it.
How often should I review clients?
A monthly routine can fit retainers; milestone reviews can fit projects. Choose a cadence that reveals a problem early enough to change scope or commitments, and compare equivalent periods.